Optimization of the Selection of Hidden Particles in the SHiP Experiment

Although the Standard Model (SM) is one of the biggest achievements in physics, it cannot explain several outstanding phenomena. This requires the introduction of new mechanisms or particles to the SM, such as Heavy Neutral Leptons or Dark Photons of the Hidden Sector. Heavy Neutral Leptons are hypothetical massive neutrino-like particles that do not couple to any SM forces, but do mix with SM neutrinos. Dark Photons are massive vector-like theorized particles that can mix with SM photons. These Hidden Sector particles are some of the biggest prospects at the SHiP experiment, which will try to discover them through the direct observation of at least two decays to the SM. This requires several background veto systems. In this thesis we suggest several optimized background veto criteria for the SHiP experiment using the kinematic properties of the reconstructed particles, regarding Heavy Neutral Leptons ("N") in the mass range between 0.7 and 1.4 "GeV/c^2", and Dark Photons ("A' ") in the mass range between 0.021 and 4.4 "GeV/c^2". The most relevant decay modes are considered, and both a cut-based approach and machine learning methods are applied. The utilization of neural networks provided the best results, with selection efficiencies above 97% for the "N \rightarrow \mu^\mp \pi^\pm" and "A' \rightarrow \mu^- \mu^+" samples. Selection studies are also shown for an alternative scenario where the Decay Vessel is set at atmospheric pressure instead of the currently planned vacuum. HS selection efficiencies and background rejection are compared between both scenarios.